New guarantees for approximating attention, accelerating SGD, and testing sample quality in near-linear time

I was a Fulbright Scholar from Turkey. It. has the goal of: "improving intercultural relations, cultural diplomacy, and intercultural competence between the people of the United States and other countries..." I guess no more. en.wikipedia.org/wiki/Fulbright…

Caltech's "Probability in High Dimensions" by Prof. Joel A. Tropp PDF: tropp.caltech.edu/notes/Tro21-Pr…

Happy to announce "Performative Prediction on Games and Mechanism Design" was accepted at AISTATS Conference 2025, and got spotlight at HAIC(ICLR 2026 workshop) with Mehrnaz Mofakhami Fernando P. Santos Gauthier Gidel Simon Lacoste-Julien (Mila and UvA) arxiv.org/abs/2408.05146 Details below 1/9🧵

Excited to be at #ICLR2025 next week! I'm currently on the job market for Research Scientist positions, especially in generative modeling, synthetic data, diffusion models, or responsible AI. Feel free to reach out if you have any openings!

Come chat with us @ ICLR on April 24th poster session to learn more about what matters in sparse LLM pretraining!

Check out our recent work on understanding Sharpness-Aware Minimization (SAM). We address several key questions regarding the convergence properties of SAM in non-convex settings. To be presented at #ICLR2025 Joint work with Dimitris Oikonomou Johns Hopkins Data Science and AI Institute,Johns Hopkins Engineering 🧵👇

This week I'll be at #ICLR25. If you like fundamental optimization results, I'll be presenting our work on surrogate losses for non-convex-concave min-max problems and learning value functions in deep RL (VIs more generally). Poster: #377 Thursday April 24 10am-12:30pm

Presenting our work, “Mastering Task Arithmetic: τJp as a Key Indicator for Weight Disentanglement,” this Friday, Apr 25, 3:00–5:30 p.m. Interested in task arithmetic? Please stop by our poster! #ICLR25 Mila - Institut québécois d'IA

I’m at ICLR presenting 2 posters. Check them out if you’re interested in deep learning theory, feature learning, and a theoretical approach to solving robustness! 📅 Thurs. 10am-12:30pm. Hall 3 + 2B, # 436“Robust Feature Learning for Multi-Index Models in High Dimensions” [1/2]

If you are at ICLR 2026 and are interested in making your RLHF really fast come find Michael Noukhovitch, used to be @ICLR 2025 and me at poster #582.

Today is my last day at Google. I started over 8 years ago, with a mandate to build a team doing bleeding edge AI research from Montreal, in what would be the first big tech AI research lab in the city. These years led to countless amazing scientific contributions from my team,

Happy to share that Compositional Risk Minimization has been accepted at #ICML2025 📌Extensive theoretical analysis along with a practical approach for extrapolating classifiers to novel compositions! 📜 arxiv.org/abs/2410.06303

Deliberate practice is accepted to #ICML2025 as a spotlight (top 2.6%!) 🚀

It's strange to think it would be at all controversial to post this. But in a time of moral decay, the obvious becomes controversial.

I'm delighted to share that our paper has been accepted by #TMLR! We empirically observed signs of scaling laws regarding how the choice of pre-trained models affects OOD test errors and Expected Calibration Error on downstream tasks.

My thesis is now online! umontreal.scholaris.ca/items/f8670d1c… This is more than just a list of publications. I invested a lot of time and passion writing this thesis in hope that it will make for an interesting read. Here's a summary of what you'll find in it.

Presenting CRM at #ICML2025 📌 Wednesday, 16th July, 11 am 📍East Exhibition Hall A-B (E-2101) Lets chat about distribution shifts! Been deep into causality & invariance based perspectives, and recently exploring robust LLM pretraining architectures.

On my way to ICCOPT I decided to give PEPit (github.com/PerformanceEst…) a try, and I wish I had used it sooner. In just a few hours I was able to confirm our theoretical results in our recent paper and I was able to get intuition that originally took me months without using it. 1/N

I’m also excited to be presenting this work (openreview.net/forum?id=4ZX2a…) at ICCOPT at USC. Theory aside there are some applications that may interest ppl in RL, games, and performative prediction. Let me know if you are in the area and want to chat!